Hugging Face's logo

Datasets:
PLAN-Lab
/
mTSBench

Modalities:
Tabular
Text
Time-series
Formats:
csv
Size:
10M - 100M
ArXiv:
Tags:
timeseries
multivariate-time-series
anomaly-detection
model-selection
benchmark
Libraries:
Datasets
Dask
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
Dataset Viewer
Auto-converted to Parquet Duplicate
timestamp
stringdate
2005-07-24 00:00:00
2005-08-13 23:30:00
in_count
int64
0
61
out_count
int64
0
40
is_anomaly
int64
0
0
2005-07-24 00:00:00
0
0
0
2005-07-24 00:30:00
0
1
0
2005-07-24 01:00:00
0
0
0
2005-07-24 01:30:00
0
0
0
2005-07-24 02:00:00
0
0
0
2005-07-24 02:30:00
0
2
0
2005-07-24 03:00:00
0
0
0
2005-07-24 03:30:00
0
0
0
2005-07-24 04:00:00
0
0
0
2005-07-24 04:30:00
0
0
0
2005-07-24 05:00:00
0
0
0
2005-07-24 05:30:00
0
0
0
2005-07-24 06:00:00
0
0
0
2005-07-24 06:30:00
0
0
0
2005-07-24 07:00:00
0
0
0
2005-07-24 07:30:00
3
2
0
2005-07-24 08:00:00
0
0
0
2005-07-24 08:30:00
0
0
0
2005-07-24 09:00:00
1
0
0
2005-07-24 09:30:00
1
1
0
2005-07-24 10:00:00
0
0
0
2005-07-24 10:30:00
0
0
0
2005-07-24 11:00:00
0
1
0
2005-07-24 11:30:00
0
0
0
2005-07-24 12:00:00
0
0
0
2005-07-24 12:30:00
0
0
0
2005-07-24 13:00:00
0
0
0
2005-07-24 13:30:00
0
0
0
2005-07-24 14:00:00
0
0
0
2005-07-24 14:30:00
0
0
0
2005-07-24 15:00:00
0
0
0
2005-07-24 15:30:00
2
0
0
2005-07-24 16:00:00
0
0
0
2005-07-24 16:30:00
1
0
0
2005-07-24 17:00:00
1
1
0
2005-07-24 17:30:00
1
2
0
2005-07-24 18:00:00
4
4
0
2005-07-24 18:30:00
0
1
0
2005-07-24 19:00:00
0
2
0
2005-07-24 19:30:00
2
1
0
2005-07-24 20:00:00
0
0
0
2005-07-24 20:30:00
0
0
0
2005-07-24 21:00:00
0
3
0
2005-07-24 21:30:00
0
2
0
2005-07-24 22:00:00
0
0
0
2005-07-24 22:30:00
0
0
0
2005-07-24 23:00:00
0
2
0
2005-07-24 23:30:00
0
0
0
2005-07-25 00:00:00
0
1
0
2005-07-25 00:30:00
0
0
0
2005-07-25 01:00:00
0
0
0
2005-07-25 01:30:00
0
0
0
2005-07-25 02:00:00
0
0
0
2005-07-25 02:30:00
0
0
0
2005-07-25 03:00:00
0
0
0
2005-07-25 03:30:00
0
0
0
2005-07-25 04:00:00
0
0
0
2005-07-25 04:30:00
0
0
0
2005-07-25 05:00:00
0
2
0
2005-07-25 05:30:00
2
2
0
2005-07-25 06:00:00
2
1
0
2005-07-25 06:30:00
1
1
0
2005-07-25 07:00:00
2
0
0
2005-07-25 07:30:00
0
0
0
2005-07-25 08:00:00
0
0
0
2005-07-25 08:30:00
7
2
0
2005-07-25 09:00:00
6
1
0
2005-07-25 09:30:00
13
5
0
2005-07-25 10:00:00
16
4
0
2005-07-25 10:30:00
5
9
0
2005-07-25 11:00:00
3
1
0
2005-07-25 11:30:00
5
7
0
2005-07-25 12:00:00
9
11
0
2005-07-25 12:30:00
2
4
0
2005-07-25 13:00:00
12
7
0
2005-07-25 13:30:00
18
3
0
2005-07-25 14:00:00
10
14
0
2005-07-25 14:30:00
7
7
0
2005-07-25 15:00:00
17
10
0
2005-07-25 15:30:00
10
6
0
2005-07-25 16:00:00
8
13
0
2005-07-25 16:30:00
5
8
0
2005-07-25 17:00:00
0
9
0
2005-07-25 17:30:00
3
3
0
2005-07-25 18:00:00
1
6
0
2005-07-25 18:30:00
0
5
0
2005-07-25 19:00:00
4
2
0
2005-07-25 19:30:00
3
5
0
2005-07-25 20:00:00
0
1
0
2005-07-25 20:30:00
0
3
0
2005-07-25 21:00:00
3
0
0
2005-07-25 21:30:00
2
0
0
2005-07-25 22:00:00
0
1
0
2005-07-25 22:30:00
0
2
0
2005-07-25 23:00:00
0
3
0
2005-07-25 23:30:00
0
3
0
2005-07-26 00:00:00
0
0
0
2005-07-26 00:30:00
0
4
0
2005-07-26 01:00:00
0
0
0
2005-07-26 01:30:00
0
2
0
End of preview. Expand in Data Studio

mTSBench

mTSBench is a collection of 344 multivariate time series from 19 datasets commonly used in anomaly detection research. Each folder corresponds to one dataset and contains *_train.csv, *_test.csv, and *_val.csv files. See data_summary.csv for per-file statistics.

How to download

This repository uses Git LFS for the CSV files.

git lfs install
git clone https://huggingface.co/datasets/PLAN-Lab/mTSBench

Load with Hugging Face

Select one of the 19 configurations so files with different schemas are not combined.

from datasets import load_dataset

calit2 = load_dataset("PLAN-Lab/mTSBench", "CalIt2")

df_train = calit2["train"].to_pandas()
df_test = calit2["test"].to_pandas()

Each CSV contains a timestamp column, dataset-specific feature columns, and a binary is_anomaly label.

Dataset overview

Dataset Domain #TS #Dims Length #AnomPts #AnomSeqs
CalIt2 Smart Building 1 3 >5K 0 21
CreditCard Finance / Fraud Detection 1 30 >100K 219 10
Daphnet Healthcare 26 10 >50K 0 1–16
Exathlon Cloud Computing 30 21 >50K 0–4 0–6
GECCO Water Quality Monitoring 1 10 >50K 0 37
GHL Industrial Process 14 17 >100K 0 1–4
Genesis Industrial Automation 1 19 >5K 0 2
GutenTAG Synthetic Benchmark 30 21 >10K 0 1–3
MITDB Healthcare 47 3 >500K 0 1–720
MSL Spacecraft Telemetry 26 56 >5K 0 1–3
OPPORTUNITY Human Activity Recognition 13 33 >25K 0 1
Occupancy Smart Building 2 6 >5K 1–3 9–13
PSM IT Infrastructure 1 27 >50K 0 39
SMAP Spacecraft Telemetry 48 26 >5K 0 1–3
SMD IT Infrastructure 18 39 >10K 0 4–24
SVDB Healthcare 78 3 >100K 0 2–678
CIC-IDS-2017 Cybersecurity 5 73 >100K 0–8656 0–2546
Metro Transportation 1 6 >10K 20 5
SWAN-SF Industrial Process 1 39 >50K 5233 1382

Citation 

@article{zhou2026mtsbench,
  title={mTSBench: Benchmarking Multivariate Time Series Anomaly Detection and Model Selection at Scale},
  author={Zhou, Xiaona and Brif, Constantin and Lourentzou, Ismini},
  journal={Transactions on Machine Learning Research},
  year={2026}
}

Links

Downloads last month
1,003
Number of rows:
24,249,172
Total file size:
4.15 GB

Paper for PLAN-Lab/mTSBench